As a broadband network is popularized on a large scale, a type of an internet service increases with great rapid, the quality of a service is continuously improved and a bandwidth requirement of a user increases rapidly, a broadband access network of a telecom operator faces an increasing load pressure, and a bandwidth capacity expansion speed of the operator cannot follow a demand growth speed of a broadband user gradually. Meanwhile, since diversities of habits and rest time and so on are more and more obvious, a traditional mode of expanding the capacity according to the number of newly-added users is high in cost and cannot meet a network development requirement.
The development of a broadband network needs to depend on sharing large-scale resources to effectively utilize resources, instead of blind physical bandwidth resources. In a traditional broadband network, network bandwidth is planned according to the number of users, so as to share bandwidth resources in cell-level and small-local-spot level, including line bandwidth resources and service node bandwidth resources. In view of a current network development speed, compared with the service node bandwidth resources, the line bandwidth resources are relatively sufficient, and the service node bandwidth resources have obviously encountered a great resource trouble. A representative problem refers to that most of users cannot enjoy a maximum bandwidth at an evening peak period, wherein the maximum bandwidth is promised by the operator. The reasons are that: most of the users are centralized in residential districts of a city at this period of time, the utilization rate of bandwidth resources of service nodes of the residential districts is extremely high, and service nodes of office districts of the city are commonly at a low use rate or even under an idle state. Conversely, the resource utilization rate of broadband network devices in the residential districts of the city within a working period is much lower than that within an evening peak period.
How to solve the problem of resource utilization efficiency of a broadband network so as to utilize existing network resources to the greatest extent, to reduce a frequency of expanding a capacity and to reduce a network cost input, which is one of main problems confronted by global telecom operators currently. In the field of broadband access, the improvement of the resource utilization rate of a service node (namely, a Broadband Access Server (BAS), a Service Router (SR) and other devices) is particularly taken as a main requirement. A mainstream technical solution in the current art is a pooling solution of service nodes, all-weather broadband resource can be reasonably utilized in large-scale regions by means of large-scale service node resource sharing, and peak-valley resource complementation in different regions and resource sharing between different types of users are achieved under limited resources.
In terms of a technical solution for resource pooling of service nodes, resource sharing of user information and network information between different service nodes is achieved firstly, a current mainstream solution refers to a distribution negotiation between the service nodes so as to achieve information all-sharing between the service nodes in a Full Mesh mode. However, as a resource sharing scale gets larger, this mode will inevitably cause the problems of huge quantity of connections and difficult in management, the change of full-network configurations will be inevitably caused by adding and deleting a service node, and the problems of difficult expandability and high management difficulty exist.
Meanwhile, as the type of user service is rich increasingly, requirements on different grades of quality guarantee for different services become more important, and the operator can not only meet requirements on provision of differentiated service for users, but also save network resources.
An effective solution has not been proposed for the problems, in the related art, of difficult expansion, difficult management and the like in a solution of sharing network resources.